This invention relates to fingerprint identification devices and, more particularly, to a compact fingerprint identification device which produces a light image of fingerprint ridge patterns upon a dark-field background.
The field of biometrics involves identifying people by measuring parts of their bodies. Many automated techniques are currently in use or under development, including fingerprint, palm print, finger pores, hand geometry, iris, retina, and face recognition. Fingerprint identification is one of the oldest and best-established methods of objective identification, originally being used in law enforcement. It now promises to find wide acceptance as a convenient and secure alternative to typed passwords, mechanical keys, or written signatures for access to computers, facilities or vehicles, and identification for financial transactions.
There are a number of different identification tasks, from determining the identity of an unknown person based on a fragment of the pattern of an unknown finger, to verifying the identity of a known person to a level of certainty, based on the pattern of a specific finger. The pattern acquisition system requirements vary accordingly, with high resolution, field of view and contrast allowing more certain identification. The FBI has published a set of standards for law enforcement fingerprint recording, the Automatic Fingerprint Identification System (AFIS) standard. The AFIS standard provides a guideline for the design of a high-quality fingerprint acquisition system.
Human fingertips have distinctive patterns of curved ridges, with a period of about 0.5 mm, depth of about 0.1 mm. Finger tissue scatters red light with a diffuse reflectivity of about 50%, and the refractive index of a finger as measured with an Abbe refractometer is about 1.51. It is desirable to have as large a field of view as possible with minimum distortion to provide more features for identification and more margin of error in finger placement.
Ridge patterns are recorded, and software extracts the coordinates and classes of features like ridge ends and bifurcations (called xe2x80x9cminutiaexe2x80x9d). There is also a line of tiny pores on the ridges that is more difficult to resolve, but can be used to provide more information. For a finger in air, ridges may be seen by the specular reflection of light from a localized source, but image contrast is limited by the underlying scattering, and tipping of the finger so it is not perfectly flat on the imaging surface. The rounded shape of the finger can cause unacceptable distortion of the image. With contact methods, the user flattens the fingertip against a surface; then ridges and valleys can be distinguished by height. There are electronic sensors that measure capacitance variation, and optical sensors that view the finger pressed against a transparent platen or window. Optical contact sensors record changes of specular reflectance, imaged onto a sensor such as a CCD or CMOS detector array.
Most fingerprint identification devices are bright-field devices, that is, they produce a dark fingerprint ridge patterns on a light background. To produce a fingerprint image with acceptable contrast, additional optical components are required to generate an uniformly bright background. Because of additional components, it is difficult to make a compact bright-field device.
U.S. Pat. No. 5,900,993 issued May 4, 1999 and entitled xe2x80x9cLens Systems for Use In Fingerprint Detectionxe2x80x9d describes a lens system having a first and second lens in combination with a third cylindrical lens to reduce optical distortion and further describes the problems associated with using holographic optical elements, off-axis optics, and tilting the object plane.
An approach using cylindrical lenses requires additional components and inherently complicates the alignment of the lens system because a lack of symmetry causes the alignment process to handle an extra degree of freedom in lens placement.
What is needed is a compact fingerprint identification device having high quality images with minimum distortion which can be adapted for use in a small volume such as an in a keyboard or an electrical key or switch, and which contains a minimum number of components and to facilitate production.
The present invention is advantageously adapted to provide a compact, low cost fingerprint imaging apparatus with minimal image distortion. The apparatus has a platen with a finger receiving surface, mounted in a housing and includes a light source mounted in the housing providing radiation which is scattered from a finger receiving surface along an optical axis. An image sensor, with rectangular pixels or square pixels combined to form a rectangle, is used to receive the fingerprint image. The optical system has a first lens disposed between said finger receiving surface of the platen and the image sensor along the optical axis, a second lens disposed between the first lens and the image sensor along the optical axis such that the first lens and second lens form an afocal optical system which images a square area on the tilted platen finger receiving surface onto the rectangular pixels of the image sensor to minimize anamorphic, non symmetric distortion. An aperture stop is placed between the first lens and the second lens to block radiation from the light source so the blocked radiation does not contribute to the image focused on the plurality of rectangular pixels.
It is therefore an object of the present invention to provide a fingerprint imaging apparatus with a minimum number of components having a minimum volume with minimum depth and width such that the apparatus is compact enough to fit within a computer keyboard, mouse-like devices and point of sale terminals.
It is another object of the present invention to provide a mechanical design with plastic components which facilitates manufacturing a low cost device.
It is a further object of the present invention to provide a fingerprint imaging apparatus with an image sensor with rectangular pixels product that will allow convenient acquisition of finger ridge images electronically; interfacing with software that allows enrollment and identification.